Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: RE: Replacing missing values

From   "joe J." <>
Subject   RE: st: RE: Replacing missing values
Date   Thu, 22 Apr 2004 11:12:47 +0000

Thanks Nick. True, linear interpolation is not a great idea. Another option is to regress investment series on related variables (eg. fuel) and time trend and use the fitted values to replace missing values. But I think this would be okay if I am dealing with capital stock rather than investment.

In fact I toyed with the idea of cubic interpolation, and wonder if the fluctuating character of investment decisions is a justification for using it?


From: "Nick Cox" <>
To: <>
Subject: st: RE: Replacing missing values
Date: Thu, 22 Apr 2004 11:22:02 +0100

You have first to consider the possibility
that your data do _not_ lend themselves
to linear interpolation. If, for example,
investment varies irregularly, as seems
plausible given that is a response to both
macroeconomic and microeconomic factors, then
gaps in a very peaky time series will be difficult to
interpolate accurately.

That said, so long as all your data are
positive, interpolating on the logarithms and
exponentiating the result is a sure
mathematical way of avoiding negative predictions.

Real economists will be able to add more. My
economics education stopped with A-level (British
high school leaving qualification) Economics
in 1969.


joe J.

> I have a panel (unbalanced) dataset of manufacturing plants
> for several
> years. Investment series is missing for a lot of plants. I tried the
> statacode -ipolate-, which uses linear interpolation and
> extrapolation
> procedures. But these have resulted in many negative values.
> Is there any
> way one could restrict -ipolate- from extrapolating negative
> values or are
> there superior methods to replace missing values?

*   For searches and help try:
Pay Cash on delivery of  products. Free Registration on

* For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index